Data & Decisions · 10 min read
Earned Value in Plain English: Know If a Project Is Really on Track
Spending 45% of the budget means nothing until you know how much work it bought. Four numbers and two ratios tell you if a project is on track.
Dennis AlejoBusiness consultant · Project & operations manager · · 2,262 words
Imagine you are managing a building project with a budget of one million dollars. In the monthly meeting, someone proudly announces that you have spent 45% of the budget and that you are "right on track." Is that good news? It depends entirely on one question nobody has asked yet: how much of the building exists? If 45% of the building is finished, you are on track. If only 30% is finished, you are in trouble, and you will not find out until the money runs low. If 60% is finished, you are doing brilliantly.
Spending alone tells you nothing. Schedule alone tells you nothing either. A project can be on schedule because everyone is working overtime, and over budget because of it. What you need is a way to put cost, time and progress on the same page, in the same unit. That is what earned value management does, and despite its intimidating name, it is a surprisingly simple idea that anyone can use, from a contractor managing a road to a founder managing a launch.
This article explains it in plain English, with one worked example, a calculator you can try below, and an honest look at where it works well and where it does not.
The road trip analogy
Picture a road trip with a fuel budget. You are driving 1,000 kilometers and you have budgeted a certain amount of money for fuel. By noon your plan says you should have covered 400 kilometers. Now check the dashboard at noon. You have actually driven 350 kilometers, and you have spent 45% of the fuel money.
Two separate questions come out of that. First, are you behind schedule? You planned 400 and did 350, so yes, a little. Second, are you efficient with fuel? You have covered 35% of the distance but spent 45% of the fuel money, so you are burning more than planned for the distance covered. Notice that "how much fuel have we used" cannot answer either question alone. Only by comparing progress with both plan and cost do you see the full picture. Earned value is that comparison, formalized.
Four numbers, in plain English
Earned value uses four basic quantities. Learn these and you have 80% of the method.
- Budget at completion (BAC): the total approved budget for the work. In our example, one million dollars.
- Planned value (PV): the value of the work you planned to have finished by today. If the plan says 40% should be done, PV is 40% of the budget, or $400,000.
- Earned value (EV): the value of the work you have actually finished by today, measured in budget terms. If 35% is truly done, EV is $350,000. This is the "earned" part: you earn value by finishing work, not by spending money.
- Actual cost (AC): what you have actually spent so far. Say $450,000.
Read them as a sentence: "We planned to have $400,000 of work done. We have actually completed $350,000 of work. It cost us $450,000." Everything else is arithmetic on those four.
Two questions, two indexes
From the four numbers come two simple ratios, each answering one question.
The Cost Performance Index (CPI) is earned value divided by actual cost. It asks: for every dollar we spend, how many dollars of progress do we get? In the example, $350,000 divided by $450,000 gives 0.78. You are getting about 78 cents of progress for every dollar. A CPI above 1.0 means you are under budget for the work done, and below 1.0 means over budget.
The Schedule Performance Index (SPI) is earned value divided by planned value. It asks: how much of the planned work did we actually complete? Here $350,000 divided by $400,000 gives 0.875. You have done about 88% of the work that the plan expected by now. Above 1.0 means ahead, below means behind.
You may also see the same ideas as differences instead of ratios: cost variance is EV minus AC (here minus $100,000) and schedule variance is EV minus PV (minus $50,000). Ratios are better for comparing projects of different sizes, and differences are better for talking about money in a meeting. Use whichever your audience prefers.
From today's numbers to a forecast
The most useful thing earned value gives you is not a diagnosis but a forecast. If the project keeps performing as it has so far, how much will it cost at the end? The classic simple formula divides the total budget by the cost index. In our example, one million divided by 0.78 gives about $1,286,000. That is the estimate at completion, and it says you are heading toward an overrun of roughly $286,000 unless something changes.
Please treat that number as a starting point for a conversation, not as a prophecy. It assumes the future will resemble the past. Sometimes that is wrong in your favor: perhaps the early work was the hard part, or perhaps a one-time problem has been solved. Sometimes it is wrong against you: later phases often cost more per unit than early ones. The value of the forecast is that it forces the question "why would the rest be different?" and demands a better answer than hope.
There is one more useful number if you want it: the efficiency you would need from now on to finish on budget. In the example you have $650,000 of work left and $550,000 of budget left, which means you would need to get $1.18 of progress for every dollar from here on. If your track record is 78 cents, that gap is the whole discussion.
The hard part is measuring progress honestly
The arithmetic fits on a napkin. The difficult part is the "35% complete" figure, because it is where optimism hides. Everyone has seen the task that stays "90% done" for weeks. Three habits keep progress honest.
First, define progress by outputs, not effort. A task is not 50% done because people have worked on it for half the planned time. It is done when a specific, checkable thing exists. Break the work into pieces small enough that "done" is obvious: a wall is poured, a report is approved, a feature is released.
Second, use simple crediting rules for each piece. Many teams use a zero-or-hundred rule for short tasks: it earns nothing until it is finished and then earns everything. For longer tasks, use milestones, for example 25% when the materials are on site, 50% when installation starts, 90% when it is installed and 100% when it passes inspection. The point is to make the credit depend on events, not on feelings.
Third, let someone who is not the doer confirm the progress. A site supervisor, a client representative or a peer. A pattern I trust from managing contracts on large construction programs is that progress reports hold up best when the person claiming progress and the person verifying it are different people. I wrote about the same idea in how a clinic opens on time, where the person who accepts the work should be the person who will use it.
Try the calculator below with the numbers from your own project. Enter the total budget, how much of the work the plan expected to be done by now, how much is truly done and what it has cost so far.
When earned value works, and when it does not
Earned value works best when the scope is reasonably defined and the work is physical or at least countable. Construction, manufacturing, infrastructure, equipment installation and many implementation projects fit well. You can say what the finished thing is, break it down and measure it.
It works less well when the scope is discovered along the way. In a research or exploratory project, or in a software product where learning changes the plan every two weeks, "percent complete" is slippery because the denominator keeps moving. For those, it is usually better to track throughput, outcomes delivered and cost per period, and to use earned value only for the parts with a stable scope. I discuss how to choose the right style of management in agile, waterfall or hybrid.
There is also a known blind spot in the schedule index. Because planned value and earned value both converge on the total budget as the project nears its end, SPI drifts toward 1.0 even for a project that is badly late. So in the last third of a project, ask for a more direct measure as well: how many calendar days or weeks behind are we on the key milestones? Do not let a friendly ratio hide a late finish.
A simple version for small projects
You do not need specialized software. For a small project, make a table with one row per piece of work: its budget, its planned completion by now, its actual completion and its cost so far. Multiply budget by percent done to get earned value, add the columns and divide. Color the two indexes green when they are 0.95 or above, amber between 0.85 and 0.95, and red below. Review it weekly or monthly, depending on the pace of the project.
That is all a small contractor, a restaurant opening team or a four-person software group needs. The numbers will be rough at first, and that is fine. What matters is the habit of asking, every week, "how much did we really finish and what did it really cost?" It connects naturally with the idea of choosing a small set of numbers that drive decisions, which I describe in data-driven project management.
Reading the signals
The two indexes together tell a story that neither tells alone.
- Over budget and behind (both below 1.0): the classic warning. The project is slow and expensive. Stop and look for the root cause: unclear scope, wrong resources, rework.
- Under budget but behind (CPI above 1.0, SPI below 1.0): you are efficient but slow. This often means too few people or waiting on something. The fix may be more capacity rather than more discipline.
- Over budget but ahead (CPI below 1.0, SPI above 1.0): you are buying speed. That may be a deliberate choice, but make sure someone chose it.
- Under budget and ahead (both above 1.0): good news, but check that quality is not being sacrificed and that progress is honestly measured.
A second example: opening a small lodge
To see that the idea is not only for big construction, imagine a team fitting out a six-bungalow garden lodge, a project not unlike Comtury Lodge. The fit-out budget is $300,000 and the plan has ten work packages: furniture, kitchens, bathrooms, lighting, landscaping, signage, systems, supplies, staff training and the final inspection. Four weeks before opening, the plan says 70% of the value should be in place. The team honestly credits what is done: furniture and lighting are complete, kitchens are installed but not accepted, landscaping is half done, training has not started. After applying the crediting rules, earned value comes to 58% of the budget. Spending so far is 66%.
The two indexes now say something sharp: CPI is about 0.88 and SPI is about 0.83. Two things stand out. The money is going out faster than the progress, and the unstarted training is a schedule risk that no cost report would have revealed. The manager can now make concrete decisions: bring training forward, pause the landscaping extras and ask the contractor why the kitchens have not been accepted. Without the numbers, the team would have felt "mostly ready" right up to the opening week, which is exactly the trap described in how a clinic opens on time.
This example is a made-up illustration, not data from a real project, but the pattern is real: earned value pulls hidden problems into the open while there is still time to fix them.
Common mistakes
A few errors appear again and again. The first is changing the baseline quietly. If the plan is revised every time reality diverges, the indexes always look fine and tell you nothing. When scope genuinely changes, record the change formally, with its cost and time impact, and keep the original baseline for comparison. I go deeper on that in how to stop scope creep.
The second mistake is counting materials or deposits as progress. Money spent on materials that sit in a warehouse is cost, not earned value. The third is measuring too rarely. A monthly look at a three-month project is a post-mortem, not control. The fourth is reporting numbers without action. If the indexes say you are in trouble and the plan does not change, the exercise was theater.
Making it a habit
Put the four numbers and the two indexes on the same page you use for your weekly or monthly review. Add one sentence from the project owner explaining the single biggest reason for any gap and one decision to correct it. Keep a short log of those decisions. After a few cycles, you will see patterns, and the project will become easier to predict.
If you manage work for others, for example as an owner's representative watching a contractor, this method also gives you an objective language for conversations about payments and delays. It replaces "I feel they are behind" with "their schedule index is 0.82 on the structural package," and that changes the tone of the whole relationship. That is part of what I do in the operations and owner's representation service, and the same discipline runs through Project Focus.
How does your team currently decide that a task is really finished? Share your rule in the comments.
Frequently asked questions
What is earned value management in simple terms?
It is a way to compare what you planned to do, what you actually did and what it cost, all in the same unit of money. With those three views you can tell whether a project is over or under budget and ahead or behind schedule, and forecast where it will end.
What do CPI and SPI mean?
CPI, the cost performance index, is earned value divided by actual cost and shows how much progress you get per dollar spent. SPI, the schedule performance index, is earned value divided by planned value and shows how much of the planned work you completed. Above 1.0 is good, below 1.0 is a warning.
How do I measure percent complete honestly?
Define progress by finished outputs instead of effort, give each task a simple crediting rule such as zero-or-hundred or milestone-based, and have someone other than the person doing the work confirm it.
Can earned value be used in agile or exploratory projects?
Only partly. It works for stable-scope pieces. For work where learning keeps changing the plan, track throughput, outcomes delivered and cost per period, and use earned value as a conversation starter rather than a verdict.
Join the conversation
Did this help? A quick reaction tells me what to write next.
Be the first to react.
Loading comments…